2.0 Analysis 2.1 Non-deployment of Anchor Even if the boatswain was not necessarily in the immediate area at the time orders were given to let go the starboard anchor, the chief officer, after ensuring that it was safe to drop the anchor, could have returned to the windlass and executed the order, albeit with a delay of a few seconds. As both the pilot and the master were on the bridge at the time the vessel's port anchor had previously been weighed, both should have been aware of the state of readiness of the anchor for immediate deployment. As immediate deployment of an anchor is a pre-requisite for handling emergency situations in confined spaces, both the master and the pilot should have ensured that the anchor was ready and the station appropriately manned, especially since berthing was to take place in stern wind conditions and without the tugs secured to thevessel. The time at which orders should have been given to let go the starboard anchor would vary depending on the assessment and realization of the developing situation. If the engine did not respond appropriately and in a timely manner, the orders to let go the anchor should have been given when this was realized. The pilot's normal practice was to initiate the turn to starboard when the vessel was a ship's length away from the knuckle of Pier 23. However, on this occasion, the first astern movement had reportedly been given earlier owing to the strong stern winds, but the vessel barely started swinging to starboard when the bow was in line with the knuckle of Pier 23. Hence, the need for deployment of the anchor should have been realized well before the vessel reached the knuckle. In such a case, the anchor could have been deployed. However, if the engine responded promptly and appropriately, the order to drop the anchor should have been given when it was realized that the headway could not be stopped in time, which would be later in the manoeuvre. The evidence from the ship's personnel and the pilot is consistent with their respective positions. Thus, in the absence of an entry in the bridge movement book of the time that let go starboard anchor was ordered, the precise timing of the order in relation to the complete manoeuvre cannot be established nor is it possible to ascertain the precise impact that the deployment of the anchor would have had on the outcome of the occurrence. As the pilot had the conduct of the vessel and the master was monitoring the vessel'sprogress, both had to have witnessed the situation developing. In any event, the anchor was not deployed in a timely manner, and the NIRJA struck the HAMILTONENERGY because she did not successfully negotiate the turn to starboard nor was she stopped in the available distance. 2.2 Speed of Approach There was conflicting evidence regarding the speed of the vessel during approach. On the one hand, the pilot maintained that the master wanted him to go faster. Since he considered this request to be reasonable, he explained that he would go from dead slow ahead to slow ahead for some two minutes (1240 to 1242) and then revert to dead slow ahead. On the other hand, the master indicated to the pilot that he considered the vessel's speed to be considerable for the conditions. In this case, any increase in speed would obviously have exacerbated the situation. However, it is known that the vessel proceeded on dead slow ahead for the major portion of the seven minutes (1242 to 1249) before the main engine was stopped prior to the astern movement. It should be noted that the manoeuvring characteristics of the vessel show that a consistent dead slow ahead movement can give a speed of 4 to 5kn. As the master's preference was at variance with the pilot's approach, the master was well aware of the limitations/difficulties of the manoeuvre. Consequently, it is unlikely that the master would have requested the pilot to proceed faster. Of the nine minutes before the striking, the main engine was stopped for the first five minutes, between 1249 and 1254. During the following four minutes, between 1254 and 1258, astern movements were ordered, the last three minutes of which (between 1255 and 1258) for full astern, followed by emergency full astern. The pilot provided contradictory evidence with respect to the vessel's position when full astern was ordered. While he indicated that it was given when the vessel was some two ship's lengths from the knuckle of Pier 23, the position he indicated on the chart placed the vessel one ship's length away; the latter is consistent with his normal approach practice. He also indicated that he had asked for full astern at least three times whereas the ship recorded that slow astern and half astern had been ordered before full astern. The NIRJA nonetheless struck the tanker. This would suggest that either the vessel's engine did not respond in a timely manner or that the vessel's headway was such that, despite timely and appropriate engine response, the vessel could not be stopped within the available distance. 2.2.1 Timeliness of Main Engine Response According to the pilot, the engine response time was about 10 seconds from stop to slow astern, and another 15 to 20 seconds to attain 90rpm astern. This is consistent with the engine operating characteristics in that it takes the engine some 25 to 35 seconds to reach full astern (90rpm) from stop. If it is accepted that the engine performed satisfactorily, the engine was effectively going astern for a total of about 3.5minutes. Within this period, the engine could have been producing maximum astern rpm for less than three minutes. According to the pilot, as the vessel travelled a distance of some 1,330m in about 25 minutes, her average speed was about 1.75kn between 1230 and 1255. The tug master's observation placed the vessel some 60m from the HAMILTON ENERGY when wash from the propeller and smoke emitting from the vessel's stack were seen. This would suggest that the vessel covered a distance of about 60m in some three minutes, giving her an average speed of about 0.6kn. This corroborates both the pilot's and the master's assessments that the vessel was barely making headway at the time of the striking, and is consistent with the damage sustained by the HAMILTONENERGY. However, as the three tugs were positioned abreast of hatch No.1, a considerable distance from the vessel's stern, some time had to have elapsed between the time the engine went astern and the wash of the propeller was observed from the tug. Hence no conclusion can be drawn from this observation to determine at what time the vessel's engine responded to the astern order(s). In any event, the strong wind acting on the vessel's starboard quarter in conjunction with the three tugs pushing abreast of hatch No.1 (forward) would account for the resultant motion being maintained at the time of the striking. 2.3 Pilot/Master Relationship and Safety As indicated by the master, his preferred approach to the berthing manoeuvre would have been to secure the tugs for optimal performance and to use the knuckle to pivot the vessel. He had discussed this with the pilot. However, the pilot opted, instead, to use his normal approach of making a direct turn into the slip without securing the tugs to the vessel. As the vessel had advanced too far into the berthing manoeuvre and only the pilot could communicate directly with the tugs, the master was reluctant to take over the conduct of the vessel, particularly since his preferred approach was at variance with that of the pilot. It is well known that masters are reluctant to take over the conduct of the vessel from pilots. This occurrence once again highlights the need for both the master and the pilot to concur on the overall berthing procedure/manoeuvre before commencing such a manoeuvre. The planned manoeuvre must take into consideration, among other things, the need for additional safeguards when berthing in adverse weather conditions. Such an agreement is a pre-requisite for the safe conduct of a vessel. 2.4 Berthing Manoeuvre Considerations As the depth of the water was 8.79m and the deepest draught of the vessel was 7.29m, the vessel had an underkeel clearance of 1.5m. As such, the underkeel clearance is not considered to have been a factor in this occurrence. A review of the chart indicates that the width of the channel in the vicinity of the slip is about 240m and following the turn, at slip 23, it is about 83m. The beam of the vessel is23m and that of the tugs is about 6m. The tugs could have been secured ready for immediate release without compromising their safety. As the tugs were not secured, their assistance was limited to pushing. Consequently, the tugs were not deployed for optimal performance. Although the pilot was aware of the stern wind conditions and the limitations imposed by not securing the tugs, his planned manoeuvre did not call for the deployment of the starboard anchor or for the use of the unfendered knuckle to assist in negotiating the vessel'sturn. Instead, he elected to berth the vessel without this assistance. The consequences of possible main-engine-related operational problems were not fully appreciated. To counteract the effect of the vessel's head swinging to starboard after the main engine had been stopped, the vessel's helm, under the direction of the pilot, had been maintained at hard-a-port. The bringing of the helm to amidships earlier, before the astern movement was requested, could have initiated the vessel's swing to starboard; however, this was not done at an early stage. Further, as the tugs had not been secured to the vessel, they could not be used effectively to retard the vessel's forward motion. Since the anchor was not deployed in a timely manner, it could not be used to accelerate the vessel's swing to starboard and to retard/control the vessel's forward motion at a critical time in the manoeuvre. Thus, the pilot was left with little option but to rely on the optimal performance of the main engine and the steering in conjunction with his own optimal performance, leaving no room for error. This, however, proved unsuccessful, which would suggest that there was no contingency plan taking all pertinent factors into consideration to respond effectively to an emergencysituation. 2.5 Effect of Pilot's Rest on Performance At the time of the striking, the pilot was on his third assignment in a 24-hour period; the previous two had been trans-lake assignments. The pilot departed Port Weller at approximately 1530 the day before the occurrence on his first assignment, and his opportunity to sleep on that passage would have occurred during a period of maximum wakefulness. Any sleep obtained during this time probably would not have been restorative. Notwithstanding the requirement for a 12-hour rest period following a trans-lake assignment, the pilot boarded another vessel between 0200 and 0300, and his earliest opportunity to go to sleep would have been some time between 0400 and 0500, which is near the end of the period when he would likely obtain restful sleep. On each of these two trans-lake passages, the pilot reportedly had five hours of sleep, but he had to wake up at Mid-Lake Ontario CIP to make a report to the VTC. In view of the above, it is unlikely that the pilot had achieved five consecutive hours of restorative sleep and, consequently, he was likely accruing a sleep debt. Manoeuvres such as those carried out by the NIRJA are demanding and require high levels of concentration, good judgement, close monitoring and immediate response to a developing situation. The pilot was familiar with the NIRJA and her manoeuvring characteristics, he had extensive experience which included docking vessels in stronger wind conditions than on this occasion, and three tugs were assisting in manoeuvring the vessel, but he was unable to turn the NIRJA successfully or to avert the striking. This would suggest that the pilot, who had the conduct of the vessel, misjudged the developing situation and did not take effective action in ample time. Judgemental errors of this type are typical of the performance of a fatigued person. As the pilot had been on duty continuously for the last 22 hours, it is most likely that more than 24 hours had passed since his last significant sleep episode. Further, after 18 hours of wakefulness, a 30percent decrement in the performance of cognitive tasks can be expected. Although the pilot had obtained some sleep during his trans-lake assignments, it is probable that he was operating with a sleep debt which could compromise his ability to perform precise monitoring, control and decision-making tasks. Since the pilot's condition was not assessed on site immediately following the accident, the extent to which his performance may have been degraded by sleep debt cannot be established. However, degradation of the pilot'sperformance, as manifested in impaired judgement, probably contributed to theoccurrence. As people are poor judges of their own fatigue and alertness levels, it is unlikely that the pilot's self-assessment in accepting a second and a third assignment was objective, more so as continuing to work involved personal financial benefit. The negative effects of shift work and irregular work schedules can be mitigated by a combination of sleep schedules and control of one's environment and diet. In this instance, the pilot had control over sleep but limited control of his environment and diet. The fact that the pilot did not avail himself of a 12-hour rest period after each trans-lake assignment, but instead opted to carry out two trans-lake assignments followed by the berthing of the NIRJA within a period of 22 hours without restorative sleep, suggests that he did not fully appreciate the negative effects of shift work and irregular work schedules on his performance. 2.6 Pilot Assignments and Safety As the Seaway navigation season draws to an end and vessels attempt to clear the Seaway, additional demands are placed on pilots. This is reflected in the pilot having been given, and carrying out, three assignments over a 22-hour period and having been asked to take on another (fourth) local assignment after he had been involved in the striking. While the collective agreement allows off-duty pilots to volunteer their services and take on additional assignments, nothing in the rules allows on-duty pilots to volunteer their services and take on additional assignments. On-duty pilots cannot take on additional assignments (following a trans-lake assignment) without having 12hours of rest, except in extenuating circumstances. Nonetheless, evidence indicates that, following a trans-lake assignment, the dispatcher may allocate another assignment if no other pilot is available. However, before taking on the assignment, it is incumbent upon the pilot to ensure that he is well rested, and able and willing to carry out the additional assignment. The pilot maintained that he was well rested. In the absence of any training program to help pilots better appreciate the negative effects of their work environment, the decision to take on additional assignments can, in part, be influenced by other considerations, such as personal financial benefit, to the detriment of vessel safety. 3.0 Conclusions 3.1 Findings The vessel attempted to berth in strong stern wind conditions. The vessel did not successfully negotiate the turn to starboard and her approach speed was such that she was not stopped in the available distance. The anchor was not deployed in sufficient time to ensure its desired effect; i.e., to successfully negotiate the turn or avert the striking. The tugs were not effectively deployed for optimal performance and could not be used to retard the forward motion of the vessel at a critical point in the manoeuvre. At a critical stage in the vessel's manoeuvre, the pilot was limited to the use of the main engine to retard/control the vessel's forward motion. It is possible that the vessel's engine did not respond to the full astern movement(s) in a timely manner or that the orders were given too late to stop the vessel before the striking. The cumulative effect of the strong stern winds and of the tugs pushing at the forward end of the vessel would account for the resultant motion at the time of the striking. Following two trans-lake assignments, the NIRJA was the pilot's third continuous assignment allocated by the dispatcher. The pilot was required to rest for 12 hours following a trans-lake assignment, but he did not. The pilot had been on duty for some 22 hours without restorative sleep and this probably adversely affected his performance. The demands to clear large numbers of vessels from the Seaway system before its closure was, in part, responsible for the pilot being given additional assignments. Research indicates that persons suffering from sleep debt may not be capable of objectively assessing the negative effect of their irregular work schedules on their performance. There is neither a directive nor an educational program in place to provide guidance to pilots regarding work-related stress, fatigue and performance maximization associated with work schedules/environment. Although the master raised some concerns regarding the manoeuvre, he relied on the pilot's local knowledge and experience. 3.2 Causes The NIRJA, while manoeuvring in strong wind conditions under the conduct of a pilot, did not successfully negotiate the turn into the slip and struck the HAMILTON ENERGY because the vessel was not stopped in the available distance. The fact that the tugs were not secured to the vessel, that the anchor was not deployed and that the performance of the pilot was probably less than optimal contributed to the accident. 4.0 Safety Action 4.1 Action Required 4.1.1 Pilot Assignments The Board has two specific concerns arising out of this occurrence. First, current pilotage assignment practices permit extended duty days such that significant performance degradation can occur. Secondly, both the Great Lakes Pilotage Authority and the pilots themselves apparently do not fully appreciate the negative effects of fatigue on performance and the strategies for mitigating those effects. In the marine transportation industry, work schedules and working conditions are often conducive to fatigue-induced degradation of performance. Implicitly, the pilots' collective agreement does recognize the need for a fixed period of rest following trans-lake assignments. However, as evidenced by this occurrence, this rest requirement is not rigorously applied. It is understood that the American Pilotage Authority does enforce a 10-hour bridge-to-bridge(9) rest period following trans-lake assignments. The Board recognizes that seasonal peaks in Seaway traffic, such as at the close of the navigation season, create increased demand for pilotage services. Flexibility in scheduling is provided by allowing pilots to take extra assignments on their scheduled days off. However, the current practice for allocating extra pilotage assignments puts the onus on the pilots to decide if they are sufficiently rested to accept the additional work. Yet, studies have shown that sleep-deprived individuals tend to underestimate their level of fatigue. Further, the Board believes that pilots are being put in the difficult position of making a safe decision, while standing to gain personal financial benefit from accepting the extra assignment. Given the vulnerability of individuals in safety-sensitive positions to significant errors in judgement when fatigued, and given the potential consequences of such errors, the Board believes that mandatory rest provisions should be strictly enforced in the assignment of marine pilots. Therefore, the Board recommends that: The Department of Transport and the Great Lakes Pilotage Authority implement a policy and procedures for allocating pilotage assignments, such that pilots receive sufficient rest to minimize the adverse effects of fatigue on performance. The Board recognizes that strict enforcement of mandatory rest periods will not in itself ensure that no pilot will suffer the adverse effects of fatigue. Many factors beyond pilot scheduling can affect a pilot's performance while on duty. Many of these are controllable by the pilots themselves in terms of personal lifestyle modifications; e.g. off-duty activities, eating and drinking habits, sleep scheduling and sleep environment, exercise, etc. Significant literature is available to assist employees in developing personal strategies for coping with the natural physiological effects of shift work, irregular work schedules, circadian disrythmia, or extended duty hours. To assist pilots in coping with the natural stresses of operating in a 24-7 industry, the Board recommends that: The Great Lakes Pilotage Authority develop and implement an awareness program to provide guidance to dispatching staff and pilots on reducing the adverse effects of fatigue on job performance.